BEAM DYNAMICS STUDIES OF THE HIE-ISOLDE TRANSFER LINES INTHE PRESENCE OF MAGNETIC STRAY FIELDS

J. Mertens∗, J. Bauche, M. A. Fraser, B. Goddard, R. Ostojić, J. S. SchmidtCERN, Geneva, Switzerland

AbstractThe ISOLDE facility at CERN produces radioactive iso-

topes far from stability for fundamental nuclear physics re-search. The radioactive beams are accelerated to high-energyusing a post-accelerator before being transferred for studyin different experiments at the end of a network of High En-ergy Beam Transfer (HEBT) lines. In the framework of theHIE-ISOLDE project, the energy of post-accelerated beamsis to be increased to over 10 MeV/u and new experimentaldetectors are being proposed for installation to exploit thenew energy regime. The stray magnetic fields associatedwith many of the new detectors will distort the beam trajec-tories in the HEBT, potentially affecting the transmissionof the low intensity beams delivered to the experiments. Inthis contribution, the influence on the HEBT of the strayfield of the proposed ISOL Solenoidal Spectrometer (ISS)is discussed, correction schemes described and shieldingoptions assessed.

INTRODUCTIONSince the completion of the first stage of the HIE-ISOLDE

project, post-accelerated beams with energies of up to5.5 MeV/u have been made available and delivered withan atomic mass-over-charge ratio of up to A/q = 4.3 totwo experimental beam lines XT01 and XT02. In the nextstage a third beam line is being installed and the XT02 lineextended to accommodate the new ISOL Solenoid Spec-trometer, see Fig. 1. In the scope of this paper, the influenceof the stray field of the ISS magnet along the third HEBT

Figure 1: A 3D visual of the HIE-ISOLDE linac, composedof three cryomodules, and the three experimental stations inthe HEBT. At the end of the first beam line, XT01, Miniballis located, at the second, XT02, the ISOL Solenoidal Spec-trometer, ISS, and the third, XT03, is reserved for movablesetups. In the drawing, XT03 is occupied by the scatteringchamber, SEC, which is presently connected to XT02. [1]

∗ jennifer.mertens@cern.ch

line as a function of beam energy and A/q was investigated.Earlier approximate studies showed that the beam transfercan be ensured for stray magnetic fields of less than 5G atthe location of the beam lines [2]. Latest 3D simulations ofthe solenoid using the coil geometry provided by the manu-facturer with its optimised magnetic shielding are shown inFig. 2.

Figure 2: 3D simulations of the solenoid’s magnetic fieldwith the latest magnetic shielding design. In addition, theenvelope of the 5G limit around the ISS magnet is shownin light red.The shielding design has been modified with two ferro-

magnetic plates separated by a gap, which makes it lighterand further attenuates the field along the beam axis, withthe main objective being to contain a 100G limit within theenvelope for personnel safety. The envelope of the 5G limitaround the ISS magnet at XT02 extends to approximately5m longitudinally and 4m transversely. XT01 and XT03are at 5.24m and 5.04m distance from XT02, respectively.Hence, they are located outside the area, where the ISS straymagnetic field exceeds 5G. The beam delivery to XT03in the case of no magnetic shielding was investigated asa worst-case scenario and to understand the impact of theoperation of the ISS on the adjacent beam line.

OPTICS AND MAGNETIC MODELThe latest beam line design from the post-accelerator up

to the end of XT03 was implemented in MADX [3]. Thecorresponding optical functions and the respective beamsizes based on the emittances given in [4] along XT03 arepresented in Fig. 3, where the dashed lines indicate the min-imum available aperture given by the quadrupoles.

The magnetic field was exported from the 3D field simula-tion every 10 cm throughout the HEBT lines. The integratedtransverse field components (Bx, By) were translated into

MOPIK102 Proceedings of IPAC2017, Copenhagen, Denmark

ISBN 978-3-95450-182-3768Co

pyrig

ht©

2017

CC-B

Y-3.

0an

dby

ther

espe

ctiv

eaut

hors

05 Beam Dynamics and Electromagnetic FieldsD01 Beam Optics - Lattices, Correction Schemes, Transport

Figure 3: Optics and beam sizes along XT03.

thin lens dipole kicks (αx,y), where possible, in-betweenexisting equipment such that an equivalent integrated kickwas imparted to the beam:

αx,y =

∫By,x

(Bρ)dl (1)

with Bρ denoting beam rigidity.

EFFECTS ON BEAM TRANSFERThree different effects relevant to the operation of the

accelerator were studied, each for various kinetic beam en-ergies ranging from 0.3MeV/u to 5.9MeV/u:

• First, the nominal transfer throughout XT03 with andwithout the presence of the ISS stray magnetic field, e.g.if the ISS is turned off due to a quench or maintenanceat the experiment, was investigated.

• The second study takes into account that the intensityof most radioactive isotope beams is so low that theyare effectively invisible to the available instrumentationin the transfer line. Thus, the setting-up is done usinga pilot beam of a nearby A/q, which is observable.Then the LINAC and HEBT are scaled linearly to thetarget A/q. However, changing the A/q also changesthe beam rigidity, which then can be more susceptibleto the stray magnetic field.

• Finally, a comparison of the required steering strengthsfor compensating the stray magnetic field to the steer-ing induced by random misalignment of the HEBTquadrupoles was done.

Ramping the ISS MagnetThe effect on the beam trajectory along XT03 due to

turning the ISS magnet on and off is shown in Fig. 4. TheRMS of beam positions measured at the 10 BeamDiagnostic

Figure 4: Energy dependent Y RMS measured on BPMsbefore and after correction as well as after turning the ISSmagnet off.

Boxes (BDB) stays below 230 µm for all investigated beamenergies for the case of the ISS magnet being on and no cor-rection being applied. A maximum excursion of 1.1mm isobserved in the first quadrupole of the last triplet at an energyof 0.3MeV/u, which corresponds to 0.18σ, and the beammoves a maximum 110 µm at the location of the experiment(0.08σ). At an energy of 5.5MeV/u the maximum excur-sion is 0.08σ and the offset at the target 0.03σ. Correctingthe trajectory using the MADX built-in MICADO and SVDtechniques results in an RMS of below 30 µm along the line.The correction counteracts the stray field very well. There-fore, when the ISS magnet is ramped down and turned offthe trajectory inverts in almost exactly the same amount asin the case with the solenoid on without correction.

A/q Scaling StudyTo study the effect of scaling the HEBT the stray field was

taken into account and the trajectory corrected for A/q = 4.5.Then the A/q of the beam was changed and the HEBT scaledaccordingly, down to A/q = 2.0. The trajectory is perturbedbecause the stray field does not scale with the A/q of thebeam, whereas the rigidity of the beam does. The results

Figure 5: Energy dependent Y RMS measured on BPMsafter correction for the ISS magnetic stray field at A/q = 4.5and various scaling of A/q.

Proceedings of IPAC2017, Copenhagen, Denmark MOPIK102

05 Beam Dynamics and Electromagnetic FieldsD01 Beam Optics - Lattices, Correction Schemes, Transport

ISBN 978-3-95450-182-3769 Co

pyrig

ht©

2017

CC-B

Y-3.

0an

dby

ther

espe

ctiv

eaut

hors

are presented in Fig. 5. It was found that even for a verylarge scaling of A/q = 2, the trajectory RMS does not exceed300 µm. Amaximum excursion of 1.3mm is observed in thefirst quadrupole of the last triplet for A/q = 2 at an energyof 0.3MeV/u, which corresponds to 0.22σ, and the beammoves at maximum 150 µm at the location of the experiment(0.11σ). At an energy of 5.5MeV/u themaximum excursionis 0.10σ and the offset at the target 0.05σ.

Misalignment StudiesTwo cases were investigated, a gaussian distributed mis-

alignment of σy = 0.1mm and one of σy = 0.2mm [5].The distorted trajectories with and without the ISS magneticstray field were corrected for 200 error seeds. Figure 6 showsthe resulting corrector settings for seven correctors from thepost-accelerator to the experimental station on XT03 forthe case of σy = 0.1mm misalignment at the lowest beamenergy. The effect of the stray field manifests itself as asystematic shift in the corrector strengths required to correctfor both the stray field and quadrupole misalignments. Therequired deflection angles in all cases are below 0.8mrad atan energy of 0.3MeV/u, which is well within the specifica-tions of the correctors [6]. The effect is less pronounced forhigher energies.

Figure 6: Corrector settings along XT00 and XT03 for σy =

0.1 mm misalignment of the quadrupoles with and withoutthe presence of the ISSmagnetic stray field for a beam energyof 0.3MeV/u.

The contribution of the quadrupole misalignment to thetrajectory distortion dominates the contribution of the mag-netic stray field, as can be seen from Fig. 7. Nevertheless,misalignment is static, whereas the magnet status (on/off)can change during operation. As already shown in Fig. 4 theeffect of ramping the ISS magnet on the trajectory RMS isat most around 230 µm, independent of misalignment.

In addition, the influence of misalignment of the ISS mag-net was investigated. The additional calculated trajectoryexcursion in X due to the off-axis field, however, would bestill two orders of magnitude smaller than any excursionshown above.

CONCLUSIONIn order to ensure stable operation of the HEBT lines the

influence of the ISS magnetic stray field and quadrupolemisalignment on the beam transport along the XT03 linehas been investigated. It was found that the stray field does

(a)

(b)

Figure 7: Energy dependent Y RMS measured on BPMsover all seeds for (a) σy = 0.1mm and (b) σy = 0.2mmmisalignment of the quadrupoles in XT00 and XT03.

influence the beam trajectory, but the maximum expectedexcursion as well as the offset at the experimental station aresmaller than a quarter of the beam size at those locations.A move of 0.11σ at the experiment is not negligible, butthis is only an issue at low energy and can be controlledby procedures, e.g. ensuring no change of the ISS magnetstatus during low energy beam operation. The reduction ofthe stray field due to the latest shielding design still remainsto be checked.

REFERENCES[1] Y. Kadi et al., “Post-accelerated beams at ISOLDE”, J. Phys.

G: Nucl. Part. Phys., JPhysG-101803.R1, 2017.

[2] M. A. Fraser, “Limits for StrayMagnetic Fields of Experimentsat HIE-ISOLDE”, CERN EDMS 1289865, 2015.

[3] MADX website: http://cern.ch/madx

[4] M. A. Fraser, “Beam Dynamics Studies of the ISOLDE Post-accelerator for the High Intensity and Energy Upgrade”, PhDThesis, University of Manchester, 2012.

[5] A. Behrens et al., “Presentation of survey results for smoothingof XL, XT00, XT01 and XT02”, CERN EDMS 1710342, 2016.

[6] A. Parfenova et al., “HIE-ISOLDE HEBT beam optics studieswith MADX”, CERN-ACCNOTE-2014-0021, 2014.

MOPIK102 Proceedings of IPAC2017, Copenhagen, Denmark

ISBN 978-3-95450-182-3770Co

pyrig

ht©

2017

CC-B

Y-3.

0an

dby

ther

espe

ctiv

eaut

hors

05 Beam Dynamics and Electromagnetic FieldsD01 Beam Optics - Lattices, Correction Schemes, Transport